Apple won’t give up control of the iPhone

Happy Sunday. It’s a busy week ahead for tech, with lots of companies reporting Q4 earnings, a big tech CEO hearing in DC, the Vision Pro going on sale, and more.

In this issue: The intent of the EU’s Digital Markets Act, which goes into effect on March 7th, is to loosen the control of tech’s gatekeepers. Apple has now shown that it has no intention of letting that happen — regulation be damned. I believe the company is on the wrong side of history here, and that karma will come back to bite it.

First, a programming note: I’ll be interviewing Pinterest CEO Bill Ready on Monday morning at the first-ever Common Sense Summit in San Francisco. Other speakers include Sam Altman, Hillary Clinton, and United States Surgeon General Vivek Murthy. If you’re reading this and will be there, please say hi.

An archive of past issues is also available online. If you aren’t already subscribed, sign up here to get future issues in your inbox.

Apple executives rarely, if ever, give on the record interviews. When they do, the conversations are fairly benign and usually centered on a product announcement.

With that in mind, Phil Schiller’s decision to speak with Bloomberg News about Apple’s stance on the EU’s Digital Markets Act (DMA) caught my eye. Beyond Apple’s press release, those two paragraphs are the only on the record quotes I’ve seen explaining the company’s view now that the DMA is soon going into effect:

“Apple is having to create technology to allow an app to install other apps, and inherent in that is risk,” Phil Schiller, a longtime Apple marketing chief who now runs the App Store, said in an interview Thursday. “That could be a big threat vector for the privacy, security and integrity of your device. So Apple is putting in place technologies and policies to try and minimize that risk.”

Schiller added that in a big year for elections worldwide — with the associated threats of malware and malicious technologies circulating — there may be “new risks for users” from downloading apps from alternative app marketplaces that don’t share Apple’s security standards. The company will not extend its content-related guidelines to a third-party marketplace.

Once you get past the unsubstantiated fear-mongering about this election year, you realize that Schiller leaves out the fact that, under Apple’s new rules, apps distributed outside of the App Store in the EU will still have to be “notarized for security and system integrity.” Nothing is really changing here; Apple is going to keep scanning everything. It’s also going to exert control over third-party app store approval.

As Eric Seufert puts it, Apple is essentially saying “heads I win, tails you lose” to developers here. The most egregious new policy is charging anyone who opts out of the current App Store fee structure a new Core Technology Fee of 50 euro cents per app install after 1 million downloads, including updates. (For those with a short memory, developer backlash to this kind of fee structure is what recently caused a near-implosion at Unity and the ousting of CEO John Riccitiello.)

When you play with Apple’s own fee calculator, it’s clear that the company’s goal is to keep developers locked into the way things already work. Even if they manage to get approved by Apple, the most likely companies to offer alternative app stores on iOS — Meta, Google, and Microsoft — aren’t going to pay what would amount to hundreds of millions of euros per year for the privilege. The new fee structure is even worse for smaller developers looking to escape the App Store.

“Under the App Store’s new fee structure for Europe, if you make $10 million in sales, Apple’s cut is $6.2 million annually,” Nikita Bier, an entrepreneur who has sold viral apps to Facebook and Discord, posted on X. “Assuming you have no operating costs or salaries, your take home amount: $2 million after tax — or 20% of your sales… If you make less than $0.57 per user — which is most apps — you will end up negative and you will owe Apple money.”

Given that none of this is actually about user safety, and that Apple clearly has no intention to comply with the spirit of the DMA, I’m not surprised by the universally negative reactions I’ve seen. Epic Games CEO Tim Sweeney called Apple’s approach to the DMA “hot garbage.” Spotify CEO Daniel Ek: “a new low.” David Heinemeier Hansson, the creator of Ruby on Rails and CTO of 37signals: an “extortion regime.”

Ultimately, it’s the fault of EU regulators that Apple was able to contort the vagueness of the DMA to maintain the status quo. There’s still a chance that the EU can force changes (or huge fines) after the DMA takes effect in early March. But regardless of what happens next, Apple has made it clear that control over iOS app distribution is going to have to be pried from its cold, dead hands.

As someone who, like so many, became interested in technology because of Apple’s computers (over a decade ago, I worked for an Apple fan site called Cult of Mac!), the company’s stance on sideloading bums me out. I’m perplexed by the insistence to continue framing it as a security issue while the Mac seems to be doing perfectly fine. The only conclusion I can draw is that it is a disingenuous argument intended to distract from Apple’s real concern: losing control and money.

Apple is nothing like the company that Phil Schiller joined in the late 80s. It won the mobile era and has become fantastically wealthy as a result. It deserves credit for creating best-in-class software and hardware, which it charges people a premium for. It can continue that approach and have a lasting, tremendous business that continues to benefit from economies of scale.

But that’s not what is happening. Rather than embrace being a neutral platform, the company’s approach to the App Store is needlessly protectionist and punitive to the developers who help make its platform valuable. Meanwhile, a whole new generation of AI-enabled, wearable hardware is emerging, and Apple’s construct of how apps should work feels less relevant by the day.

Unfortunately, I have to agree with Ek: “This is a company resting, not breaking any new ground, and turning its back on the principles that once made it the shining example of innovation.”

If you’re interested in a deeper dive on Apple’s approach to the DMA, I recommend this analysis from John Gruber. This paragraph in particular stood out:

“The delicious irony in Apple’s not knowing if these massive, complicated proposals will be deemed DMA-compliant is that their dealings with the European Commission sound exactly like App Store developers’ dealings with Apple. Do all the work to build it first, and only then find out if it passes muster with largely inscrutable rules interpreted by faceless bureaucrats.”

My thoughts on what else is happening in tech right now:

I’ve enjoyed reading all the reactions to my interview with Mark Zuckerberg about Meta’s ambition to build artificial general intelligence. Below is some of the best commentary I’ve seen:

Jack Clark, co-founder of Anthropic, in his Import AI newsletter:

For a few years I used to sit around with colleagues working on AGI and we’d list out major tech companies and work out whether they were betting on AGI or if they weren’t, why they weren’t. For many years, Facebook was one of those peculiar companies which had an AI research lab but due to a combination of (seemingly) cultural and focus reasons wasn’t making a high-conviction bet on AGI (though was doing tons of great research). The recent dramatic rise of large language models seems to have sparked more attention into AGI at Facebook and it seems like Zuckerberg is now pivoting the company’s vast R&D budget towards AGI more directly. Thus, a well capitalized shoe has now fallen out of the sky and made contact with the earth.

Matthew Lynley in his AI industry newsletter, Supervised:

If AGI is the general-ish goal of AI developers, Meta’s Llama-series models are essentially a replay of earlier strategies: ingratiating itself into a permanent place in developer workflows at the expense of other companies, and picking off the best parts that come out of it. Meta has already shown it can do it at least twice with the AI community, with both PyTorch and with its Llama-series models sparking a focus on smaller performant models rather than a one-size-fits-practically-all foundation model like one out of OpenAI.

Casey Newton in Platformer:

When you consider how many processes Meta is running at a global scale, from recommendation algorithms to content moderation, it becomes more apparent how much it can benefit from automating and personalizing them. In that respect, continuously improving AI is the rising tide that lifts all boats.

M.G. Siegler on his new blog, Spyglass:

It all sounds fairly dystopian, but there’s undoubtedly some entertainment value here. The question is if there’s any actual value after the novelty wears off. Most of these AI chat bots seem to be fun for about five seconds, notably Meta’s faux celebrity bots. Part of the issue there is obvious: you know you’re not talking to the actual celebrity and perhaps all of this is only interesting if you at least think you are talking to an actual person.

Rob Leathern, a former trust and safety leader at Meta and Google, on his Substack:

Neither company has (yet) been forthcoming about how they plan to use powerful new large language models (“LLMs”) to slow down the spread of scams and deceptive advertising on their platforms. Google and Meta are the two biggest walled gardens in media: so it’s fair for us all to ask for details of how those walls will be defended and maintained.

Arun Rao, lead product manager on Meta’s generative AI team, on LinkedIn:

The biggest risk of being open is that more can go wrong and you leave yourself more exposed to scrutiny and criticism, but that to me is worth it.

Some interesting career moves I’ve noticed recently:

That’s it for this issue. If you aren’t already subscribed to Command Line, don’t forget to sign up and get future issues delivered directly to your inbox.

I always appreciate your feedback and story tips. Thanks for subscribing.